Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.

com

Ann. rheum. Dis. (1973), 32, 413

Articular mobility in an African population

P. BEIGHTON,* L. SOLOMON, AND C. L. SOSKOLNEt
Department of Orthopaedic Surgery, University of the Witwatersrand, South Africa

There is considerable variation in the range of move- of this investigation and to discuss the clinical signifi-
ments which are possible in the joints of normal cance of the observations which were made.
individuals. In general, females are more mobile than
males, while joint laxity decreases with age (Ellis and Methods
Bundick, 1956; Wynne-Davies, 1971). (a) Outline of survey
Ethnic differences in joint mobility have also been The project was carried out in the Tswana village of
described. For instance, Negroes and Indians have Phokeng, which is situated at the foot of the Magaliesburg
of Hills, about 60 miles
been shown to have a greater range movements The population, although living to the north-west of Johannesburg.
than Caucasians of the same age and sex (Harris and were relatively sophisticated and readily in a rural environment,
co-operated in the
Joseph, 1949). Similarly, in an investigation in Cape investigation.
Town, Indians were found to be more loose-jointed The survey had been preceded by a census, and by this
than indigenous Xhosa and Hlubi, who in turn had a means an epidemiologically valid section of the population
greater degree of joint laxity than white South was chosen for the study. Within this group, measurements
Africans (Schweitzer, 1970). ofjoint mobility were made on 1,081 individuals of all ages.
Articular mobility is a graded trait and at one end Details of the demography and investigation techniques
of the spectrum a considerable degree of joint laxity will be published elsewhere.
may occur in normal individuals (Wood, 1971). (b) Assessment ofjoint mobility
Apart from this form of hypermobility, joint laxity is Joint laxity was measured during the clinical examination
also a component of a variety of genetically deter- by means of a series of simple tests which have been used to
mined syndromes (McKusick, 1966; Beighton, 1970). assess the range of articular movements in a variety of
It can also occur in the absence of other stigmata as a hypermobility syndromes (Beighton and Horan, 1969).
simple inherited entity (Sturkie, 1941; Carter and The method was a modification of a technique which was
Sweetnam, 1958, 1960; Beighton and Horan, initially developed by Carter and Wilkinson (1964).
Patients were given a numerical score of 0 to 9, one point
1970). being allocated for the ability to perform each of the
It has been suggested that hypermobile individuals following tests, as depicted in Figs 1 to 5:
are prone to orthopaedic disorders, such as degenera-
tive joint disease, dislocations, joint effusions, and (1) Passive dorsiflexion of the little fingers beyond 900.
muscular pains (Hass and Hass, 1958; Kirk, Ansell, (2) Passive apposition of the thumbs to the flexor aspects
and Bywaters, 1967; Grahame, 1971). Articular of the forearms.
laxity has also been implicated as an important factor
in the genesis of congenital dislocation of the hip
(Wynne-Davies, 1970). For these reasons, hyper-
mobility may well be of considerable clinical signifi-
cance.
An epidemiological survey has recently been com-
pleted among the Tswana people of the Western
Transvaal. The main aims of this investigation were
the study of various bone and joint conditions, but .I.
the survey also provided an excellent opportunity for
the measurement of the range ofjoint movements in a
large number of individuals and for the assessment of
the influence of age, sex, and somatotype on their
articular mobility. The importance of joint laxity in
the production of non-specific musculo-skeletal FIG. 1 Hyperextension of thefifthfinger. In thisparticular
complaints was also evaluated. illustration, the extension angle does not reach the required
The purpose of this paper is to present the results 900
Accepted for publication December 6, 1972.
Presented to the Heberden Society in November, 1972.
* Professor of Human Genetics, Medical School, University of Cape Town, Cape Town, S. Africa.
t Research Officer, Human Sciences Research Council, Pretoria, S. Africa.
23
 Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com
414 Annals of the Rheumatic Diseases
FIG. 2 Apposition ofthe thumb to the ventral aspect of the
forearm
P -
FIG. 3 Hyperextension of the elbow joint beyond 100
(3) Hyperextension of the elbows beyond 100.
(4) Hyperextension of the knees beyond 100.
(5) Forward flexion of the trunk, with knees straight, so
that the palms of the hands rested easily on the floor.
These tests were all easy to perform and the results repre-
sented quantitative measurements.
Measurements were made of the angle to which the fifth
finger of each hand could be passively extended. The
patient's right or left handedness was also noted.
The findings were recorded on standardized profor-
mata, coded, and statistically analysed by computer. The
results were expressed as 'fitted curves'. These curves were
determined by the method of 'least squares', a statistical
procedure which minimizes the sum of the squares of the
vertical distances of the plotted points to a trend line. This
technique is of value for indicating trends in data.
(c) Assessment of arthralgic symptoms
To facilitate the evaluation of their relationship with joint
mobility, musculoskeletal symptoms were graded 0 to 4.
One point was scored for an affirmative answer to any of
the following questions:
(i) Any pains in the hands or feet?
(ii) Any other joint pains?
FIG. 4 Hyperextension of the knee joint beyond 10°
(iii) Any backache?
(iv) Any other pains in the limbs?
This method of assessment of arthralgic symptoms was
relatively crude, but in view of the language barrier and
the circumstances of the survey, it was the best that could
be achieved. However, as all the individuals who were
examined were graded in the same way, it is reasonable to
assume that they could be compared on the basis of these
observations.
(d) Assessment of somatotype
The relationship between joint laxity and body build was
examined by correlating the mobility score with the
ponderal index (PI) and metacarpal length.
The PI is an expression of somatotype which is derived
by dividing the height in inches by the cube root of the
weight in pounds (Acheson and Chan, 1969). This index
was calculated for 101 patients in the survey in the 20 to
28-year age group, in whom mobility scores on the 0 to 9
scale were also available.
Skeletal configuration can be expressed in terms of the
length of the 2nd metacarpal. The distance between the
proximal and distal articular surfaces of the left 2nd meta-
carpal were measured for 94 individuals in the 20 to 28-
year age group, using antero-posterior radiographs of the
 Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com

Articular mobility in an African population 415

were aged 20 or over. The percentage frequency distri-

butions of the mobility scores in these groups are
shown in Fig. 6. It is apparent from the fitted curves
and the actual data that 94 per cent. of the males and
80 per cent. of the females achieved scores of 0, 1, or 2.
This range of movement can therefore be regarded as
normal for adult Tswana.
The relationship between the mobility score, as an
expression of joint laxity and age, is shown in Fig. 7.
The curve indicates that at any age, females are more
mobile than males. However, in both sexes, the degree
of joint laxity diminishes rapidly throughout child-
hood, continuing to fall at a slower rate in adult life.
5s

4' -* Males (n=456)

os -----o Females (n=625)
0
3,-
0

-o 2-
0
-

0
0 20 40 60 80
FIG. 5 Placing the palms ofthe hands flat on thefloor while Age (years)
maintaining the knees in full extension FIG. 7 Relationship between mobility score (O to 9) and
age, for males andfemales
hands which had been taken at standard distances. The
values obtained were then correlated with the mobility
scores, for each of these investigation subjects. (b) Fifth finger extensibility
The angle to which the fifth finger could be extended
Results was recorded in 398 males and 537 females. The
(a) Mobility score averages of these measurements were then compared
A mobility score on the 0 to 9 scale was obtained in with average mobility scores for males and females in
1,081 individuals, of whom 168 males and 334 females four age cohorts (Table I, overleaf). The correlation
70 coefficients within these cohorts varied between 0 373
and 0-559.
60- - * Males (n=168)
When the extensibility angles and mobility scores
-0 Females (n=334) were averaged for the total patients of each sex,
u 50 correlation coefficients of 0 599 and 0-533 were ob-
I- tained for males and females respectively.
0~
c 40
D
4)
(c) Joint laxity and handedness
30- The relationship between handedness and joint move-
a- ments, as expressed by fifth finger extensibility, is
20 shown in Table IL (overleaf). These results, derived
from 299 males and 361 females, are tabulated
10- separately for each sex, in age cohorts. The extensi-
0-
* -Q___4___ O ___ O bility value is less on the dominant side for each sex in
almost every age group.
2 3 4 s 6 7 8 The occurrence ofleft-handedness in approximately
Mobility score (0-9) 5 per cent. of males and 3 per cent. of females is in
FIG. 6 Percentagefrequencydistribut 'ion of mobility score accordance with established norms for handedness in
in adults other populations.
 Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com

416 Annals of the Rheumatic Diseases

Table I Fifth finger extensibility and mobility score, by age and sex
Fifth finger SD Avergae
Sex Age group No. of extensibility Correlation No. of mobility
(yrs) patients (0) (0) coefficient patients score (0-9) SD
Male 0-19 249 75 20 0 559 288 25 2-3
20-44 69 58 23 0373 74 1P0 1-4
45-64 43 41 26 0 550 51 03 09
65+ 37 37 24 0-491 43 04 07
Total 398 65 26 0 599 456 1-8 2-1
Female 0-19 237 70 21 0-415 291 3-0 2-4
20-44 101 60 24 0 502 108 1.9 1.9
45-64 107 51 25 0 523 120 1-2 1-4
65+ 92 44 24 0-432 106 09 1-5
Total 537 60 25 0 533 625 2-1 2-2

Table II Handedness and average fifth finiger extensibility, by age and sex
Handedness
Right Left
Age group
Sex (yrs) Average fifth finger Average fifth finger
Patients extensibility (0) Patients extensibility (0)
No. Per cent. Right Left No. Per cent. Right Left
Male 0-19 167 59 75 81 10 67 78 75
(299) 20-44 55 19 54 63 4 27 88 60
45-64 33 12 41 47 0 0
65+ 29 10 35 46 1 6 20 20
Total 284* 100 1St 100
Female 0-19 154 44 72 82 11 92 71 71
(361) 20-44 72 21 58 69 0 0 0 0
45-64 72 21 48 64 0 0 0 0
65+ 51 14 45 55 1 8 10 10
Total 3491 100 12§ 100
97
I} of total females.
per

; 15 per cent. of total males. §3 per cent.

(d) Joint laxity and musculo-skeletal symptoms Discussion

The relationships between joint mobility on the 0 to
9 scale and musculo-skeletal symptoms are shown in
Table III. In any age cohort, the score for musculo-
skeletal symptoms is always positively related to the
mobility score. This relationship is most evident in
females, the overall correlation coefficients being of
the order of 0 797 for males and 0 957 for females.
(e) Joint laxity and somatotype
The correlation coefficients between the PI and the
mobility score were -0-082 for the 47 males and
+0 097 for the 54 females in whom this relationship
had been investigated. Similarly, correlation coeffi-
cients between metacarpal length and mobility scores
were -0-185 for the 46 males and +0 008 for the 48
females.
The method of assessing joint mobility by means of a
score based upon the range of movements of certain
joints has proved its value in previous investigations.
Although the great majority of normal adults score
values of 0, 1, or 2 on the mobility scale, individuals
with inherited hypermobility syndromes achieve
scores at the top end of the scale (Beighton and
Horan, 1969). It is therefore reasonable to conclude
that the scale is valid for the measurement of joint
mobility.
The results which were obtained demonstrate that
the range of movements decreases with age, falling
rapidly during childhood, and more slowly through-
out adult life. However, at any age, females are con-
sistently more hypermobile than males. These findings
are in accordance with the observations which Ellis
 Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com

Articular mobility in an African population 417

Table III Joint mobility (O to 9) and musculo-skeletal symptoms, by age and sex
Patients
Sex Age group Musculo-skeletal
(yrs) Mobility score No. Per cent. symptoms
Male (168) 20-44 0-2 65 88 0K123
3-5 8 11 05
6-9 1 1 0.0
Total 74* 100
45-64 0-2 49 96 0-959
3-5 2 4 1.000
6-9 0 0 0.0
Total 51t 100
65+ 0-2 43 100 1-512
3-5 0 0 00
6-9 0 0 00
Total 431 100
Female (334) 20-44 0-2 74 68 0-649
3-5 27 25 0 704
6-9 7 7 1.000
Total 108§ 100
45-64 0-2 100 83 1200
3-5 19 16 1-579
6-9 1 1 2-000
Total 1201 100
65+ 0-2 92 87 1-402
3-5 11 10 1-909
6-9 3 3 2'333
Total 106 T 100
* 44 per cent. of males. t 30 per cent. of males. 11 36 per cent. of females.
t 51 per cent. of males. § 32 per cent. of females. i 32 per cent. of females.

and Bundick (1956) made during their investigations
on American Negroes.
On a basis ofanatomical and radiological studies of
finger joints, Schweitzer (1970) postulated that the
factor which limited articular movements was tension
in the volar plate and flexor tendons. These structures
are largely composed of collagen which, in the skin at
least, becomes stiffer in the elderly (Ridge and Wright,
1966). In addition, the intrafibrillar cross-links in the
collagen are increased with advancing age (Verzar,
1957). The reduction in joint mobility with ageing is
readily explicable on the basis of these changes.
A positive relationship exists between the mobility
score and the angle to which the fifth finger can be
passively extended. This relationship is independent
of age and sex, and measurements of this angle may
therefore be used in the assessment of overall joint
mobility. However, it is not always an easy matter to
make accurate measurements of the fifth finger ex-
tensibility angle. In view of the simplicity of the 0 to 9
scale and of the fact that it takes movements of a wide
range ofjoints into consideration, this method is to be
preferred for the assessment of generalized articular
mobility.
The fifth finger hyperextensibility measurements
show mobility in this pair of joints to be inversely
related to left and right handedness. In other words,
the range of movements is almost invariably dimin-
ished on the dominant side. Although exact measure-
ments were not made, a strong clinical impression has
been gained that all the other paired joints conform to
this pattern.
The findings indicate that a significantly positive
relationship exists between joint laxity and arthralgic
complaints in individuals of any age and sex. This
observation supports the hypothesis of Kirk and
others (1967) who postulated that hypermobile indi-
viduals may be predisposed to musculo-skeletal
problems.
It is generally assumed that tall, thin individuals
tend to be more loose jointed than those with a short
stocky physique. For this reason, correlations were
sought between joint mobility and body build, as ex-
pressed by PI and second metacarpal length. In
 Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com

418 Annals of the Rheumatic Diseases

neither case did any significant correlation exist and This research was supported jointly by the Orthopaedic
it is therefore unlikely that there is any association Chair Trust Fund and The Carl and Emily Fuchs
between joint mobility and these facets of somatotype. Foundation.
The methodology of this survey was straight- We are indebted to the President of the Human Sciences
Research Council, Pretoria, for his kind provision of com-
forward, and the mobility measurements were easy to puter facilities, and to the secretarial and photographic
carry out. It is our intention to apply these techniques staff of the Orthopaedic Department, for their assistance
to other population groups, in order to assess ethnic in the preparation of this article. We are especially grateful
variations in articular laxity, to confirm the observed to the Tswana of Phokeng for their good-natured partici-
relationships between joint mobility, age, and sex, pation in the survey, and to Chief Mokgatle and the
and to substantiate the association between joint lax- Department of Bantu Administration for permission to
ity and musculo-skeletal problems. undertake this investigation. Mrs. Fifi Sparrow and Miss
Liz Paget typed the manuscript with enthusiasm and
Summary efficiency. P.B. was in receipt of a Geigy Fellowship
Joint mobility was measured in 1,081 members of a awarded by the Arthritis and Rheumatism Council of
Tswana community in the Western Transvaal, as Great Britain
part of an epidemiological survey of bone and joint
conditions. The method, using an assessment of the Discussion
range of movements of a predetermined set of joints, DR.R.GRAHAME(London) We have recently surveyed
was easy to carry out and gave reproducible results. the joint mobility of 300 London schoolchildren and have
Joint mobility diminished with ageing, falling rap- also found a very rapid fall-off of joint mobility between
idly as childhood progressed, and more slowly the ages of 5 and 11 years. We found no significant sex
throughout adult life. Marked differences were pre- difference and there was no difference between the domi-
sent between the sexes, females having a greater de- nant and the non-dominant sides. The extensibility of the
gree of joint laxity than males of the same age. fifth metacarpophalangeal joint was the most valuable
A positive correlation existed between the mobility and sensitive technique. The other methods recommended
score and the degree of passive hyperextension which by Carter and Wilkinson (1964) were not sensitive
could be achieved by the fifth finger. The range of enough to differentiate between the age groups that we
movements of the fifth finger was greater on the non- examined.
dominant side in both left and right handed indivi- PROF. BEIGHTON Dr. Grahame is to be congratulated
duals. on his studies. I think that our respective results are not
Arthralgic complaints were positively related to incompatible as we were looking at different things from a
different view point. Dr. Grahame was focusing on a speci-
joint laxity in both sexes. Physique, as expressed by fic joint and examining it in a scientific and sophisticated
ponderal index and metacarpal length, was not found way. We were conducting a wider investigation of a
to bear any relation to joint mobility. diffuse situation.
References
ACHESON, R. M., AND YICK-KWONG CHAN (1969) J. chron. Dis., 21, 543 (New Haven Survey of Joint Diseases:
Prediction of serum uric acid in a general population)
BEIGHTON, P. (1970) 'The Ehlers-Danlos Syndrome'. Heinemann, London
AND HORAN, F. (1969) J. Bone Jt Surg., 51B, 444 (Orthopaedic aspects of the Ehlers-Danlos syndrome)
, (1970) Ibid., 52B, 145 (Dominant inheritance in familial generalised articular hypermobility)
CARTER, C. O., AND SWEETMAN, R. (1968) Ibid., 40B, 664 (Familial joint laxity and recurrent dislocation of the patella)
S (1960) Ibid., 42B, 721 (Recurrent dislocation of the patella and of the shoulder)
AND WILKINSON, J. (1964) Ibid., 46B, 40 (Persistent joint laxity and congenital dislocation of the hip)
ELLIS, F. E., AND BUNDICK, W. R. (1956) Arch. Derm., 74, 22 (Cutaneous elasticity and hyperelasticity)
GRAHAME, R. (1971) Proc. roy. Soc. Med., 84, 692 (Joint hypermobility-clinical aspects)
HARRIS, H., AND JOSEPH, J. (1949) J. Bone Jt Surg., 31B, 547 (Variation in extension of the metacarpophalangeal and
interphalangeal joints of the thumb)
HASS, J., AND HASS, R. (1958) Ibid., 40A, 663 (Arthrochalasis multiplex congenita)
KIRK, J. A., ANSELL, B. M., AND BYWATERS, E. G. L. (1967) Ann. rheum. Dis., 26, 419 (The
hypermobility syndrome. Musculoskeletal complaints associated with generalized joint hypermobility)
McKUJSICK, V. A. (1966) 'Heritable Disorders of Connective Tissue', 3rd ed. Mosby, Saint Louis
RIDGE, M. D., AND WRIGHT, V. (1966) Gerontologia (Basel), 12, 174 (The ageing of skin: a bio-engineering approach)
SCHWEITZER, G. (1970) S. Afr. med. J., 44, 246 (Laxity of metacarpophalangeal joints of finger and interphalangeal
joint of the thumb: A comparative interracial study)
STURKIE, P. D. (1941) J. Hered., 32, 232 (Hypermobile joints in all descendants for two generations)
VERZAR, F. (1957) In 'Connective Tissue', a Symposium organized by the C.I.O.M.S., Blackwell, Oxford, ed. R. E.
Tunbridge, p. 208.
WOOD, P. H. N. (1971) Proc. roy. Soc. Med., 64, 690 (Is hypermobility a discrete entity?)
WYNNE-DAVIES, R. (1970) J. Bone Jt Surg., 52B, 704 (Acetabular dysplasia and familial joint laxity: two etiological
factors in congenital dislocation of the hip)
(1971) Proc. roy. Soc. Med., 64, 689 (Familial joint laxity)
Downloaded from http://ard.bmj.com/ on February 8, 2018 - Published by group.bmj.com

Articular mobility in an African

population.
P Beighton, L Solomon and C L Soskolne

Ann Rheum Dis1973 32: 413-418

doi: 10.1136/ard.32.5.413

Updated information and services can be found

at:
http://ard.bmj.com/content/32/5/413.citation

These include:

Email alerting Receive free email alerts when new articles cite
service this article. Sign up in the box at the top right
corner of the online article.

Notes

To request permissions go to:

http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to:

http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to:

http://group.bmj.com/subscribe/